Abstract
SIGNIFICANCE: Non-invasive optical measures for evaluating rheumatoid arthritis (RA) remain limited. Clinical scores such as the 28-joint Disease Activity Score (DAS28) and ultrasound are operator-dependent and do not directly quantify optical tissue changes. Short-wave infrared (SWIR) hyperspectral imaging (HSI) provides sensitivity to inflammation-associated spectral changes in biological tissues. AIM: The aim is to develop and evaluate a quantitative HSI soft abundance scoring (HSISAS) method for assessing treatment response in RA patients using SWIR-HSI. APPROACH: Eleven RA patients who met the 2010 American College of Rheumatology/European Alliance of Associations for Rheumatology criteria underwent SWIR-HSI (900 to 1700 nm) of the wrist joints before and after 12 weeks of biologic therapy. The HSISAS algorithm used pixel-wise reflectance spectra together with spectral correlation structure estimation and constrained energy minimization-based abundance scoring to compute intra- and inter-subject scores. These scores were statistically compared with conventional clinical indices, including DAS28, erythrocyte sedimentation rate, C-reactive protein, and power Doppler ultrasound findings. RESULTS: Both intra- and inter-HSISAS significantly increased after treatment ( p < 0.001 ) and were positively associated with DAS28 improvement. Reflectance spectra showed visible pre/post-treatment differences, particularly near 1250 nm. The abundance maps visually differentiated pre- and post-treatment states consistent with clinical improvement. CONCLUSIONS: The HSISAS framework provides a quantitative, non-contact optical measure of treatment-related spectral changes in RA. This method may offer translational potential for disease monitoring and may complement existing ultrasound and laboratory assessments.